Ghost voltage — induced voltage on a 'dead' cable next to a live circuit

The guide on the two-pole voltage tester covers step 3 of the five safety rules: proving a circuit dead with a loaded measurement. This article covers a phenomenon that can complicate that proof: ghost voltage — a voltage measured on a correctly switched-off, and in itself dead, cable, caused by an adjacent live circuit rather than by a fault in the switching-off itself.

How ghost voltage arises

Two mechanisms underlie this, often in combination:

Capacitive coupling — a switched-off conductor that runs parallel to a live circuit over a long distance (for example in the same cable tray, conduit or cable route) forms a small capacitor with that conductor. Through that capacitance, part of the voltage of the active circuit is "coupled" onto the dead conductor.
Inductive coupling — with a live circuit carrying a significant current (for example a heavily loaded supply cable), the magnetic field of that current induces a voltage in a parallel, switched-off conductor, similar to how a transformer works.

The longer the parallel run and the closer the cables lie together, the greater the coupled voltage can be — over long parallel runs (for example in a shared cable tray over tens of metres), a high-impedance meter (such as a non-contact tester or a screwdriver phase tester) can show a voltage close to the full mains voltage of the active circuit.

Why this is not a reason to distrust step 3

As the guide on the two-pole voltage tester explains, a two-pole voltage tester compliant with EN 61243-3 performs a loaded measurement with a sufficiently low internal impedance. A capacitively or inductively coupled ghost voltage can usually only supply a very limited current — as soon as the tester draws that small loading current, the measured voltage of a genuine ghost voltage typically collapses quickly to a low residual level, whereas a genuinely live circuit stays at nominal voltage under load. This distinction is precisely why a loaded measurement is prescribed and a high-impedance, unloaded measurement is not.

Why step 4 (earthing and short-circuiting) covers the underlying risk

A low, collapsing voltage under a loaded measurement rules out that the circuit is still actively at normal operating voltage, but it does not rule out that, over a longer exposure or with a change in the adjacent circuit, a relevant voltage could reappear on the dead conductor. For that reason, the five-safety-rules method includes a follow-up step after proving the circuit dead: earthing and short-circuiting the work area. An earthed and short-circuited circuit cannot build up any dangerous voltage anymore, regardless of any capacitive or inductive coupling with an adjacent circuit — this is the measure that structurally removes the ghost-voltage risk, not the voltage measurement in step 3 on its own.

Practical relevance

When working on a long cable that runs parallel for a stretch with a heavily loaded or long adjacent live circuit — for example in a shared cable tray in a plant room, or along an outdoor cable route — finding a low, collapsing voltage while proving the circuit dead is not a reason for alarm, but it is a signal not to skip the following step (earthing and short-circuiting), precisely because the cause of that residual voltage (coupling with an adjacent circuit) remains present for as long as work continues in that area.

Common mistakes

Ignoring a low, collapsing voltage while proving the circuit dead without understanding its cause — a ghost voltage that collapses under load confirms the circuit is not actively at operating voltage, but is not a licence to skip the earthing-and-short-circuiting step.
Using a non-contact tester to "confirm" or "rule out" ghost voltage — as the guide on the two-pole voltage tester explains, this instrument is unsuitable for step 3 in any case, let alone for interpreting a coupling phenomenon.
Assuming ghost voltage is always harmless — the actual current a capacitive or inductive coupling can deliver depends on the length and configuration of the parallel run; over very long runs with heavy loading on the active circuit, the coupled energy can be higher than over a short run.
Fitting the earthing-and-short-circuiting device only after the work is finished instead of right after proving dead — it is precisely the interval between proving the circuit dead and fitting the earth that is the moment a coupling phenomenon could go unnoticed.